For autonomous vehicles, the advantages and disadvantages of Lidar are very obvious. Mounting a laser on the roof allows the car to capture millions of information points per second and quickly create 3D images of the surrounding environment.
But a major problem is that laser radar components are more expensive than cars. The high-performance laser radar on Google's early cars cost $70,000; laser radar pioneer Velodyne's shorter spectrum and narrower viewing angles still cost thousands of dollars.
There are always other ways to replace laser radar in practice. For example, Tesla's cars use cheaper radars, cameras and ultrasonic sensors. But the fatal accident of Model S this summer also shows that this alternative is still far from ideal.
What automated car manufacturers really need is a cheap and reliable sensor to replace radar and camcorders. Today, Israeli startup Oryx Vision may be able to provide a suitable method.
Oryx's technology, the coherent light radar, erases the difference between radar and lidar. This type of radar, like a lidar, uses a laser to detect the road ahead; but it also has the characteristics of a radar, treating the reflected signal as a wave rather than a particle.
This type of laser is a long-wave infrared laser, and because of its frequency characteristics during operation, it is also called a terahertz (THz) laser. Because the human eye cannot recognize the light at this frequency, Oryx can use a higher power level than Lidar. At the same time, the long-wave infrared light is absorbed by water at a low rate and is rarely affected by solar radiation. Therefore, this system, like the lidar system and camera, does not fail due to heavy fog or strong direct sunlight.
Oryx's technology is relatively cost-effective because the system does not require a mechanical mirror or a series of channels to guide the laser and capture the environment. Just emit a laser beam to illuminate the road ahead of the car (Oryx doesn't specify the system's field of view, but if it doesn't have a 360-degree view of the lidar's lidar, Oryx needs to provide additional components to look at the different directions. Case).
In fact, the most critical point is the process of reflecting the beam back to the sensor, which is why no one has mastered the technology so far. In this process, a second set of optical instruments directs incident light into a large number of miniature rectified nano-antennas. This technology was spent six years researching and inventing by Oryx partner David Ben-Basat.
The incident light produces an AC response in the rectified antenna, that is, the incident light is converted into a DC signal. Rani Wellingstein, another partner at Oryx, said, "This system is a million times more sensitive than traditional lidar systems. Because antennas treat incident light as a wave, they can detect Doppler shift. The wavelength of the wave changes due to the relative motion of the wave source and the observer), so that objects moving at high speed around the road can be detected."
Each nano-antenna in the system is approximately 5 square microns; they are ultimately assembled into an integrated circuit through a thin film chip fabrication process. In this way, it can transmit signals into the machine learning system to identify objects in the scene.
Oryx currently manufactures millions of experimental nano-antennas, but it's not enough to make a sensor that ultimately outputs single-pixel information and validate it. In the next one to two years, Oryx intends to create a prototype of 300 pixels, and then develop a system of tens of thousands of pixels composed of hundreds of thousands of nanometer antennas. The ultimate goal is to create 100,000 pixels of a million nano antennas. Level equipment for use in cars.
Wellingstein said: "At present, the radar can detect a distance of 150-200 meters, but its pixels are not high enough. The lidar has a higher pixel level, but the field of view is only about 60 meters, only 30 meters under strong light." It is hoped that Oryx's coherent light radar technology will observe debris within 60 meters of the road, pedestrians within 100 meters and motorcycles within 150 meters. Such technology is a huge improvement in the performance of today's sensor systems.
If Oryx can break the performance improvement barrier and achieve mass production, it will be the best in both range and pixel. Image source: Oryx
At the same time, the cost of the Oryx system will also decrease. If the company can achieve large-scale production of its manufacturing process (of course not yet possible), perhaps manufacturing a million-level nano-antenna will only be a little more complicated than the process of manufacturing semiconductor chips. Currently, Oryx announced that it has received $1.7 million in Series A financing and has negotiated with a number of driverless car companies including Nutonomy, which launched its first unmanned taxi in Singapore.
Joe Funke, an engineer who has conducted autonomous driving research at several car startups, said: "This idea seems to work very well. A cheaper, wider field of view sensor will always have a broad application prospect because it can improve the speed of the car. But before they make the prototype, I still remain skeptical."
But don't lose hope for the laser radar too early. After receiving a $150 million investment from Ford and Baidu, the Velodyne automotive-specific laser radar may be able to increase the application of the system exponentially. At the same time, the cost of technology is falling, and Velodyne plans to reduce the latest small laser radar to $500.
In addition, solid-state lidars are under consideration, and many solid-state lidar manufacturers, including Quanergy, Innoluce, and another Israeli startup, Innoviz, are developing devices under $100. Manufacturer Osram recently announced that in 2018, they could market a solid-state lidar of less than $50 to the market, and the credit-sized lidar could identify pedestrians within 70 meters.
These systems may not be as good as Oryx envisioned, but in addition to the ultra-fast driving on the highway, they can already meet the daily driving needs. And if this can be achieved, the era of full control of autonomous vehicles with lidars may soon be over.
5v 2a usb charger,5v 2.5a usb charger,5v 2.1a micro usb charger, 5v 2a micro usb charger
Guangdong Mingxin Power Technologies Co.,Ltd. , https://www.mxpowersupply.com